in which the foraminifera developed. The Cd/Ca ratio is
therefore a tracer for nutrient concentration of the water
masses, present and past (Marchitto and Broecker 2006;
Fig. 21.11).
Cd and
13 C have similar geochemical behaviors, with the
essential difference that surface waters can exchange their
dissolved carbon dioxide with the atmosphere, while Cd is
not involved in ocean-atmosphere exchanges. In general,
there is an excellent anti-correlation between the variations
in the Cd/Ca ratio and the
13 C/
12 C of benthic foraminifera
measured in sediment cores. One notable exception is the
Southern Ocean, where the benthic foraminifera that lived
during the last ice age have very negative d
13 C, while their
concentrations in Cd are very similar to those of recent
sediments from the Holocene. Despite inter-specific differences that may have affected d
13 C reconstructions based on
benthic foraminifera (Gottschalk et al. 2016), or diagenetic
and metabolic effects influencing the incorporation of trace
metals into biogenic calcite (Marchitto and Broecker 2006;
McCorkle et al. 1995), this discrepancy between the two
indicators remains still to be explained.
Reconstructing the Dynamics of Water Masses
The tracers we have described so far allow the reconstruction
of specific physical or chemical characteristics of water
masses, but they do not convey information on their
dynamics. In this section, we do not discuss in detail tracers
related to particle transport at the ocean floor (particle size
distribution, magnetic grain size, sortable silt). However, we
will discuss two unstable radioactive tracers in the ocean: the
concentration in
14 C of benthic foraminifera and the excess
231 Pa/
230 Th ratio in sediments.
When surface waters exchange carbon dioxide with the
atmosphere, they absorb
14 C. Currently, the
14 C concentration in the surface ocean is 95% of that of the atmosphere.
When surface waters sink, they bring with them the dissolved carbon dioxide as well as the
14 C they contain. Once
they reach the bottom of the ocean, these waters are isolated
from the atmosphere, and
14 C decreases due to its own
radioactive decay, with its period of 5720 years. The oldest
waters in the northern Pacific and Indian Oceans have an
apparent age of around 800 years (see Chap. 4). The
planktonic foraminifera (that live in surface waters) and the
benthic foraminifera (that live on the ocean floor) incorporate the
14
C of the waters around them into their shells. By
comparing, at the same level of sediment, the
14 C ages of
planktonic and benthic foraminifera, we can estimate the
apparent age of the deep waters over the last *40,000
years. This apparently simple method presents in fact many
difficulties. Firstly, despite recent analytical developments,
benthic foraminifera are not always sufficiently abundant,
and it may be difficult to obtain the amount of carbonate
required for analysis from a normal-size sample, even using
the most sensitive technique, accelerator mass spectrometry.
Second, the
14 C ages of foraminifera, once in the sediment,
are very sensitive to bioturbation: the abundance of one
species shows considerable variation over time, and shells
that are found at one level may have been displaced by the
activity of burrowing animals and therefore come from a
significantly different age level than the selected one.
Finally, the atmospheric
14 C concentration has undergone
large scale changes, so much so that the difference in
14 C age
between planktonic and benthic foraminifera does not
directly reflect the residence time of the waters at depth.
Another approach to reconstruct the dynamics of
deep-water masses makes use of the geochemical behavior
of the decay chain of uranium in seawater. The isotopic
composition of dissolved uranium is constant throughout the
ocean. Two of the isotopes of uranium,
235 U and
234 U, decay
producing
231 Pa and
230
Th respectively, with an output ratio
that is constant and equal to 0.093.
231
Pa and
230 Th are very
reactive to particles sinking in the water column: they are
adsorbed on their surface and settle as sediment along with
them. However,
231 Pa is less reactive than
230 Th, so that the
residence time in seawater of dissolved
231 Pa is close to
200 years, while that of dissolved
230 Th is only thirty years
(Yu et al. 1996). The residence time of
231 Pa is close to that
of NADW in the Atlantic. Because of this, a fraction of the
dissolved
231
Pa is advected out of the North Atlantic Ocean
by NADW (about 50% in the modern ocean), while
230 Th is
unaffected and settles completely with the particles. The net
loss of
231 Pa in the water column at depths greater than or
equal to the level at which NADW flows, leads to a deficit of
231 Pa in the sediments and therefore in the
231 Pa/
230 Th ratios
below the production ratio (0.093). If the circulation of
NADW becomes slower, less
231 Pa is advected out of the
basin, and the
231 Pa/
230
Th ratio of the particles settling to the
sediment increases to values closer to the production ratio.
It should be noted that sediments also contain
231 Pa and
230 Th, present as daughter isotopes of the uranium in clays
that have reached secular equilibrium with their relevant
parents. The addition of
231 Pa and
230 Th from settling particles thus produces an excess of these two radioisotopes in
the sediment. Measuring the ratio of excess
231 Pa/
230 Th in
the sediments therefore allows the variations in the circulation of deep waters from the North Atlantic towards the
Southern Ocean to be traced, and so to detect the variability
associated with major changes in climate (McManus et al.
2004; Gherardi et al. 2009; Lippold et al. 2016). This
technique has been used with success to reconstruct the
‘strength’ of the thermohaline circulation back to *140 ky
ago (Guihou et al. 2010, 2011; Böhm et al. 2015).
21 Climate and the Evolution of the Ocean: The Paleoceanographic …
241
therefore a tracer for nutrient concentration of the water
masses, present and past (Marchitto and Broecker 2006;
Fig. 21.11).
Cd and
13 C have similar geochemical behaviors, with the
essential difference that surface waters can exchange their
dissolved carbon dioxide with the atmosphere, while Cd is
not involved in ocean-atmosphere exchanges. In general,
there is an excellent anti-correlation between the variations
in the Cd/Ca ratio and the
13 C/
12 C of benthic foraminifera
measured in sediment cores. One notable exception is the
Southern Ocean, where the benthic foraminifera that lived
during the last ice age have very negative d
13 C, while their
concentrations in Cd are very similar to those of recent
sediments from the Holocene. Despite inter-specific differences that may have affected d
13 C reconstructions based on
benthic foraminifera (Gottschalk et al. 2016), or diagenetic
and metabolic effects influencing the incorporation of trace
metals into biogenic calcite (Marchitto and Broecker 2006;
McCorkle et al. 1995), this discrepancy between the two
indicators remains still to be explained.
Reconstructing the Dynamics of Water Masses
The tracers we have described so far allow the reconstruction
of specific physical or chemical characteristics of water
masses, but they do not convey information on their
dynamics. In this section, we do not discuss in detail tracers
related to particle transport at the ocean floor (particle size
distribution, magnetic grain size, sortable silt). However, we
will discuss two unstable radioactive tracers in the ocean: the
concentration in
14 C of benthic foraminifera and the excess
231 Pa/
230 Th ratio in sediments.
When surface waters exchange carbon dioxide with the
atmosphere, they absorb
14 C. Currently, the
14 C concentration in the surface ocean is 95% of that of the atmosphere.
When surface waters sink, they bring with them the dissolved carbon dioxide as well as the
14 C they contain. Once
they reach the bottom of the ocean, these waters are isolated
from the atmosphere, and
14 C decreases due to its own
radioactive decay, with its period of 5720 years. The oldest
waters in the northern Pacific and Indian Oceans have an
apparent age of around 800 years (see Chap. 4). The
planktonic foraminifera (that live in surface waters) and the
benthic foraminifera (that live on the ocean floor) incorporate the
14
C of the waters around them into their shells. By
comparing, at the same level of sediment, the
14 C ages of
planktonic and benthic foraminifera, we can estimate the
apparent age of the deep waters over the last *40,000
years. This apparently simple method presents in fact many
difficulties. Firstly, despite recent analytical developments,
benthic foraminifera are not always sufficiently abundant,
and it may be difficult to obtain the amount of carbonate
required for analysis from a normal-size sample, even using
the most sensitive technique, accelerator mass spectrometry.
Second, the
14 C ages of foraminifera, once in the sediment,
are very sensitive to bioturbation: the abundance of one
species shows considerable variation over time, and shells
that are found at one level may have been displaced by the
activity of burrowing animals and therefore come from a
significantly different age level than the selected one.
Finally, the atmospheric
14 C concentration has undergone
large scale changes, so much so that the difference in
14 C age
between planktonic and benthic foraminifera does not
directly reflect the residence time of the waters at depth.
Another approach to reconstruct the dynamics of
deep-water masses makes use of the geochemical behavior
of the decay chain of uranium in seawater. The isotopic
composition of dissolved uranium is constant throughout the
ocean. Two of the isotopes of uranium,
235 U and
234 U, decay
producing
231 Pa and
230
Th respectively, with an output ratio
that is constant and equal to 0.093.
231
Pa and
230 Th are very
reactive to particles sinking in the water column: they are
adsorbed on their surface and settle as sediment along with
them. However,
231 Pa is less reactive than
230 Th, so that the
residence time in seawater of dissolved
231 Pa is close to
200 years, while that of dissolved
230 Th is only thirty years
(Yu et al. 1996). The residence time of
231 Pa is close to that
of NADW in the Atlantic. Because of this, a fraction of the
dissolved
231
Pa is advected out of the North Atlantic Ocean
by NADW (about 50% in the modern ocean), while
230 Th is
unaffected and settles completely with the particles. The net
loss of
231 Pa in the water column at depths greater than or
equal to the level at which NADW flows, leads to a deficit of
231 Pa in the sediments and therefore in the
231 Pa/
230 Th ratios
below the production ratio (0.093). If the circulation of
NADW becomes slower, less
231 Pa is advected out of the
basin, and the
231 Pa/
230
Th ratio of the particles settling to the
sediment increases to values closer to the production ratio.
It should be noted that sediments also contain
231 Pa and
230 Th, present as daughter isotopes of the uranium in clays
that have reached secular equilibrium with their relevant
parents. The addition of
231 Pa and
230 Th from settling particles thus produces an excess of these two radioisotopes in
the sediment. Measuring the ratio of excess
231 Pa/
230 Th in
the sediments therefore allows the variations in the circulation of deep waters from the North Atlantic towards the
Southern Ocean to be traced, and so to detect the variability
associated with major changes in climate (McManus et al.
2004; Gherardi et al. 2009; Lippold et al. 2016). This
technique has been used with success to reconstruct the
‘strength’ of the thermohaline circulation back to *140 ky
ago (Guihou et al. 2010, 2011; Böhm et al. 2015).
21 Climate and the Evolution of the Ocean: The Paleoceanographic …
241